We present effective numerical approaches for modeling unknown system from measurement data. By utilizing deep neural network (DNN), we seek to approximate the flow map of the underlying system. Once an accurate DNN model for the flow map is constructed, it serves as a predictive model for the unknown system and enables us to conduct system analysis. We demonstrate that residual network (ResNet) is particularly suitable for modeling autonomous dynamical systems. Extensions to other types of systems will be discussed, including non-autonomous systems, systems with uncertain parameters, and more importantly, systems with missing variables. Finally, we discuss deep learning of unknown partial differential equations (PDEs).